The Truth of "Computer Audiophile"

[alfe]

Is that on a 1500++ or on a 1500-- system ?

Paul's is 1500--, some say.

 

I just try to keep it on topic ...

 

For my test only the electricity bill is on 1500--

[alfe]

Thanks Alfe, will do. We a have listening session get together scheduled for next month and so may see what the group thinks of the comparison. AlexK should be there too. Its a long way from Paris (having recently been there as you know) but if you're in Sydney let us know. hey its a lot warmer here ;-)

Cheers

David

 

I think you are more close to heaven that we are:)

[Audiophile Neuroscience]

For my test only the electricity bill is on 1500--[/quote?for

 

Yeah my gryphon monoblocks are quite economical to run as they double up as heaters in winter....i never listen to music in summer unless i want a sauna .;-]

[alfe]

For my test only the electricity bill is on 1500--[/quote?for

 

Yeah my gryphon monoblocks are quite economical to run as they double up as heaters in winter....i never listen to music in summer unless i want a sauna .;-]

 

They are not listed but in summer I usually use the Vincent STP-100 hybrid mono blocks, they are slightly less coloured than transcoded DSD:)

[Miska]

Miska agrees that all else being equal, offline upsampling should sound better than "on-the-fly."

 

Well, I don't think I ever said that. Running the process with NAA makes the difference disappear if there's any.

 

What matters much more are the algorithms and the DAC used to convert it to analog...

[Miska]

I do think it softens the treble a little, warms mids and blunts transients a bit, all giving a warmer darker mellower less harsh/edgy sound but trading off a sense of pace and rhythmic drive.

 

This kind of stuff depends on the algorithms used to do the work. For example with my exaSound e28 I get much crisper transients and more detail running it upsampled to DSD256 than it gives when run with PCM...

[alfe]

This kind of stuff depends on the algorithms used to do the work. For example with my exaSound e28 I get much crisper transients and more detail running it upsampled to DSD256 than it gives when run with PCM...

 

DSD 2048 is even better the only limit is the processor but you can still borrow a computer from the NSA:)

[Miska]

DSD 2048 is even better the only limit is the processor but you can still borrow a computer from the NSA:)

 

I have already support for DSD1024, so I'm future-proofing things...

 

DSD256 is still slower than the modulator speed used by Sabre internally when run in PCM mode. So it's not about modulator rate, but about oversampling filters and modulator design.

 

There are couple of advantages with computers and doing things at player side:

1) player software can see much more "to the future" than a DAC can, so it can predict what's coming up

2) computer with asynchronous DSP clocking can decide to go back and re-process data (even multiple times) with different strategy if output doesn't look desired

[alfe]

I have already support for DSD1024, so I'm future-proofing things...

 

DSD256 is still slower than the modulator speed used by Sabre internally when run in PCM mode. So it's not about modulator rate, but about oversampling filters and modulator design.

 

But Jriver is limited to 128.

Bricasti DAC is also limited to 128.

So for David listening session 128 will be the limit.

[Audiophile Neuroscience]

But Jriver is limited to 128.

Bricasti DAC is also limited to 128.

So for David listening session 128 will be the limit.

 

Well Brian Zolner offers free upgrades for all software changes and (at cost hardware changes - he just added a bnc digital input and replaced all fuses with circuit breakers) so I have DSD 2048 covered. Now just need need that NSA computer. ;-).... and Im sure Jim will add dsd 2048. :-)

[Miska]

But Jriver is limited to 128.

Bricasti DAC is also limited to 128.

So for David listening session 128 will be the limit.

 

OK, so the DAC puts limit there, but the used algorithms are easy to change because software is easy to change. Or alternatively use some offline converter and same software.

 

Due to AD1955 DAC chips it goes through intermediate digital filters and remodulation though...

[Jud]

Well, I don't think I ever said that. Running the process with NAA makes the difference disappear if there's any.

 

What matters much more are the algorithms and the DAC used to convert it to analog...

 

Which is why I said "all else being equal" - precisely to take what you just mentioned, and any other difference, out of the equation, and leave as the *only* difference inline versus offline conversion. In other words, same algorithms, same DAC, same everything, only difference being that in one case you use the conversion/filtering/modulator algorithms offline, in the other case the same conversion/filtering/modulator algorithms on the fly.

 

I believe when I asked if you had tried your own software offline versus on the fly, what you posted here in response was that it did sound better offline, but (1) the NAA should or did make most differences disappear; and (2) you would wind up having numerous copies of files converted offline that you would then trash when the algorithms improved, so why go through that, especially considering (1)?

 

I don't think I'm making this up, though I'll ask you to forgive me for not attempting to search for it - we both have lots of posts here, so finding it would be pretty hard.

[Audiophile Neuroscience]

I think Brian uses modified stereo ADI 1955 D/A converters in a dual mono setup and femto clocks for each channel at the dac chips/ he does have a range of minimum and linear phase filters for pcm as well as dsd filters.

[alfe]

... and Im sure Jim will add dsd 2048. :-)

 

J​riverplay 2048 version:)

[Paul R]

Is that on a 1500++ or on a 1500-- system ?

Paul's is 1500--, some say.

 

I just try to keep it on topic ...

 

I try not to listen to the whispers behind my back, comes with being a free American you know.

 

Actually, if you were to count all the bits and parts around here, the total is well into five figures. The cost of my main system was dramatically reduced when we replaced $2700 speakers with a $300 set of large Advents. Technically, they do not sound as good as say, a set of Maggie 1.7s, but to me, they sound just right. Their sins are indeed, sins of ommission to me. When/if they die, I have two more sets in reserve. Man, do they image well, which is one of the things I really like.

[4est]

Darker with less details is hardly "a tube sound" unless you have been using junky tube gear. The HQPlayer transcoding is far superior to JRiver's. I did not care for it using JRiver.

 

No wild explanation my daily job is to design PUH and drives and I can't throw all the informations in a public forum, as I said before I will not touch this subject again.

Pure DSD is superior to PCM for me, but transcoded PCM by JRiver..... I tried, bought a DAC specially for that...sound darker with less details... a tube sound:), but I learned one thing with time it's just matter of taste and everyone should respect that.

[alfe]

Darker with less details is hardly "a tube sound" unless you have been using junky tube gear. The HQPlayer transcoding is far superior to JRiver's. I did not care for it using JRiver.

 

Yep I have been using junky tube gear called audio research VTM 200 and cary SLM 200:)

You mean that HQ player is an artist in colouring and Jriver is house painter:)

[EuroDriver]

EuroDriver, I found offline upsampling to be higher quality (for some of the same reasons people report the NAA sounds better), and Miska agrees that all else being equal, offline upsampling should sound better than "on-the-fly." Of course Miska makes very fine filters/modulators, so the offline upsampler must be very good as well for comparable performance. For a number of good reasons, Miska does not want to make HQPlayer an offline upsampler.

 

DSD256 files get pretty big ;-)

 

With a well isolated NAA, I suspect that offline upsampling will have little advantage of HQ Players online process. Online has the advantage of being able to change filters and modulators on whim

[Jud]

DSD256 files get pretty big ;-)

 

With a well isolated NAA, I suspect that offline upsampling will have little advantage of HQ Players online process. Online has the advantage of being able to change filters and modulators on whim

 

Those are some of the good reasons.

[Jud]

Those are some of the good reasons.

 

I should add that the two possible reasons I can think of, in spite of the negatives, to use offline conversion/filtering are:

 

(1) If the particular offline algorithms would be better than those available with inline software; and

 

(2) A related factor, which is that the time the algorithm takes to work is not nearly so tight a limitation offline, so it opens up possibilities for offline algorithms not available "on the fly."

[PeterSt]

Okay, if we can't stay on topic anyway ...

 

The way I work (how XXHighEnd works) you'd lose on the enormous overhead of the file reading, never mind be that from RAMDisk. Say roughly put : The noise coming from that supersedes by far the noise from the real time filtering.

 

Of course this is not the same as allowing the off line filtering to take 24 hours for 5 minutes of music. So if *that* implies a better filtered file which is huge and meanwhile produces some noise, then I guess the better filtering will be profound(ly audible).

[tranz]

Psssst...........(whispering)......a tube preamp stage.....ssshhhhhhhhhhh

 

Well, sarcasm aside, I did finally have a chance to try it. Some ridiculously expensive unit. But all it did was warm up the digititus a bit.

[elcorso]

Well, sarcasm aside, I did finally have a chance to try it. Some ridiculously expensive unit. But all it did was warm up the digititus a bit.

 

The sometimes cold bits appreciate the gentle heat of the tubes (Not necessarily in winter time).

 

Roch